Optimal Model And Algorithm For Emergency Logistics Vehicle Routing Problem Considering Stochastic Demand And Robustness

Every natural disasters and other emergencies happened, they will bring about inestimable lost to human society and seriously affects the social stability. Research shows that performing the rescue immediately and efficiently in the key rescue period, that’s72hours after the disaster, plays a crucial part in decreasing the disaster losses.The operating environment of emergency logistics determines that emergency logistics vehicle routing problem is more complex than general vehicle routing problem, because that it involves both the time requirements of receiving the emergency supplies and the demands with uncertainty, and in the early stages of the disaster, the emergency supplies usually is in short supply. Therefore, the key point of efficient rescue is to identify the vehicle route and the quantity of emergency supplies reasonably.This paper aims at researching the dispatching of emergent supplies in the key rescue period after the disaster. At first, it introduces the basic concepts of emergency logistics systematically and summarizes the vehicle routing problem and algorithm, which establish the theoretical foundation for the following research in this paper.Considering the characteristics of demand uncertainty in the rescue process, it sets up the optimal model for emergency logistics vehicle routing problem considering stochastic demand, and combining the practical characteristics, adopts the improved sweep assignment and combines the average distance method to turn the multiple-depot vehicle routing problem into the single-depot vehicle routing problem, then designs the detail solving procedure based on genetic algorithm and in the end proves the effectiveness of the model and the algorithms by the calculation examples. On the basis of considering stochastic demand, further considering of the uncertainty scenario, and combining the scenario analysis method and the definition of77robustness, it sets up an optimal model and algorithm for emergency logistics vehicle routing problem considering stochastic demand and robustness, and turn the multiple-depot vehicle routing problem into the single-depot vehicle routing problem adopting the method of structuring a virtual parking lot, then also solves the model based on genetic algorithm. And by comparing the optimal solution in single scenario with the robust optimal solution in multi-scenario indicates, we can find that the latter can be better in balancing the fluctuation and influence caused by different scenarios, this also shows that the optimization method for emergency logistics vehicle routing problem considering robustness can be better to reduce all kinds of uncertainties in the scheduling of emergency supplies and reach the purpose of minimizing casualties and property losses.